multi-flow transmitter with full format and rate ... · additional number of optical carriers is...

General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

Users may download and print one copy of any publication from the public portal for the purpose of private study or research.

You may not further distribute the material or use it for any profit-making activity or commercial gain

You may freely distribute the URL identifying the publication in the public portal If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from orbit.dtu.dk on: Oct 18, 2020

Multi-Flow Transmitter with Full Format and Rate Flexibility for Next GenerationNetworks

Katopodis, Vasilis; Mardoyan, Haik; Tsokos, Christos; De Felipe, David; Konczykowska, Agnieszka;Groumas, Panos; Spyropoulou, Maria; Gounaridis, Lefteris; Jenneve, Philippe; Boitier, FabienTotal number of authors:18

Published in:Journal of Lightwave Technology

Link to article, DOI:10.1109/JLT.2018.2850800

Publication date:2018

Document VersionPeer reviewed version

Link back to DTU Orbit

Citation (APA):Katopodis, V., Mardoyan, H., Tsokos, C., De Felipe, D., Konczykowska, A., Groumas, P., Spyropoulou, M.,Gounaridis, L., Jenneve, P., Boitier, F., Jorge, F., Johansen, T. K., Tienforti, M., Dupuy, J-Y., Vannucci, A., Keil,N., Avramopoulos, H., & Kouloumentas, C. (2018). Multi-Flow Transmitter with Full Format and Rate Flexibilityfor Next Generation Networks. Journal of Lightwave Technology, 36(17), 3785-3793.https://doi.org/10.1109/JLT.2018.2850800

https://doi.org/10.1109/JLT.2018.2850800

https://orbit.dtu.dk/en/publications/ce704ad9-fd5b-42b5-985b-13e7cf649c09

https://doi.org/10.1109/JLT.2018.2850800

0733-8724 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JLT.2018.2850800, Journal ofLightwave Technology

1

Abstract—We extend our proof-of-concept demonstration of a

novel multi-flow transmitter for next generation optical metro

networks. The multi-flow concept is based on the combination of

spectrum and polarization sliceability, and its implementation on

the combination of a polymer photonic integration platform with

high-speed IQ modulators. In this work, we replace the static

scheme of our previous demonstration for the definition of the

optical flows and the generation of the driving signals, and we

unveil the true potential of the transmitter in terms of

programmability and network flexibility. Using a software

defined optics (SDO) platform for the configuration of the digital

and optical parts of the transmitter, and the configuration of the

optical switch inside the node, we demonstrate operation with

flexible selection of the number and type of the optical flows, and

flexible selection of the modulation format, symbol rate, emission

wavelength and destination of each flow. We focus on 16 specific

cases accommodating 1 or 2 optical flows with modulation format

up to 64-quadrature amplitude modulation (64-QAM), and

symbol rate up to 25 Gbaud. Through transmission experiments

over 100 km of standard single-mode fiber, we validate the

possibility of the transmitter to interchange its configuration

within this range of operation cases with bit-error rate

performance below the forward error correction limit. Future

plans for transmitter miniaturization and extension of our SDO

platform in order to interface with the software defined

networking (SDN) hierarchy of true networks are also outlined.

Index Terms — multi-format multi-rate multi-flow

transmitters, elastic optical networks, polymers, photonic

integration, software-defined optics, FPGA, InP-DHBT circuits.

I. INTRODUCTION

etro network traffic has undergone more than a threefold

increase in the last five years [1], and is expected to

continue to grow due to the widespread adoption of

Internet services and applications such as cloud computing

Manuscript received March 10, 2018. This work was supported by the EU-

funded project PANTHER (Contr. No 258846). V. Katopodis, C. Tsokos, P. Groumas, M. Spyropoulou, L. Gounaridis, H.

Avramopoulos, and Ch. Kouloumentas are with the National Technical University

of Athens, Athens 15773, Greece (e-mail: [email protected]).

P. Groumas and Ch. Kouloumentas are also with Optagon Photonics, Agia

Paraskevi 15341, Athens, Greece ([email protected]).

H. Mardoyan, P. Jennevé, and F. Boitier are with Nokia Bell Labs, Nokia Paris

Saclay Route de Villejust 91620 Nozay, France (haik.mardoyan@nokia-bell-

labs.com). D. Felipe, and N. Keil, are with Fraunhofer Institute for Telecommunications,

HHI, Berlin 1058, Germany ([email protected]).

A. Konczykowska, F. Jorge and J.-Y. Dupuy are with III-V Lab, Joint lab

between Nokia Bell Labs, Thales-TRT and CEA/Leti, 1 avenue Augustin Fresnel,

91767 Palaiseau, France ([email protected]).

M. Tienforti, and A. Vannucci are with Cordon Electronics, Via S. Martino 7,

20864 Agrate Brianza (MB), Italy ([email protected]). T. K. Johansen is with Technical University of Denmark, Ørsteds Plads, 2800

Kgs. Lyngby, Denmark ([email protected]).

Fig. 1: Networking concept regarding the placement of multi-flow transmitters

at the edge switches of metro networks. The diagram outlines the specific case

of data center gateways acting as data aggregators and as edge switches in a metro network that enables data center interconnection.

and Internet of Things [2]. This trend has a major impact on

the operation of metro networks, as it imposes strict

requirements for the traffic aggregation systems and the

transmission systems at the edge switches of these networks,

including the gateways of interconnected data centers, as

shown in Fig. 1.

Current 100G transmitter products, based on the use of

dual-polarization quadrature phase-shift keying (DP-QPSK)

modulation, have started being installed at the optical

interfaces of these edge switches, and can partially alleviate

this problem. However, the need to go to flexible interfaces

with higher capacity via the use of higher-order quadrature

amplitude modulation (QAM) formats and the use of

additional number of optical carriers is already obvious [3-5].

The combination of photonic integration with software

defined networking (SDN) is considered as the most

promising way to go to this direction, enabling next generation

networks with 400 Gb/s and 1 Tb/s links and possibility for

reconfiguration of the bandwidth resources according to the

network needs [6-11]. Integrated multi-carrier transmitters

have been proposed in particular for the generation of optical

super-channels that offer the potential to add or remove

bandwidth via the adjustment of the number of modulated

optical carriers, and the selection of the modulation format and

symbol rate [12-15]. These transmitters can also support

multi-flow operation, since they have the possibility to

aggregate their total capacity in a large optical flow for

serving high traffic demands to a single destination or slice

their spectrum and distribute their total capacity among a

larger number of smaller optical flows for serving parallel

links to different destinations [15-18]. This type of spectral

slicing represents an additional degree of flexibility, which can

Edge Metro network

Edge

Edge

Wireless access

Ethernet

LAN

DC-3

DC-4

Ethernet

LAN

DC 1

SAN

Fibre ChannelWireless

access

DC-2

Edge

Fibre Channel

SAN

Multi-Flow Transmitter with Full Format and

Rate Flexibility for Next Generation Networks

V. Katopodis, H. Mardoyan, C. Tsokos, D. Felipe, A. Konczykowska, P. Groumas, M. Spyropoulou,

L. Gounaridis, P. Jennevé, F. Boitier, F. Jorge, T. K. Johansen, M. Tienforti, J.-Y. Dupuy, A.

Vannucci, N. Keil, H. Avramopoulos, and Ch. Kouloumentas

M

mailto:[email protected]










2

Fig. 2: Layout of the multi-flow transmitter based on two PolyBoards for the generation, routing and polarization handling of the optical flows. The layout

corresponds to an integrated version of the transmitter, where a 4-fold Mach-

Zehnder Modulator (MZM) array is placed between the two PolyBoards in

order to form two IQMs.

Fig. 3: Multi-flow transmitters inside an optical node connected to the client

interfaces at the digital side and the two WSS at the optical side. Each transmitter can support either single- or 2-flow operation. The SDO platform

in this work controls the number, modulation format, symbol rate, wavelength

allocation and direction of the optical flows by controlling the data processing

unit, the optical part of the transmitter and the two WSSs.

improve the network economies, increase the switching

capacity, and save on the front panel port density of digital

switches [19-23].

Recently, we introduced a novel multi-flow transmitter

concept, which can provide additional flexibility and savings,

as it combines the spectral with the polarization slicing for

reconfigurable generation of optical flows [24, 25]. Work by

other group has also evolved along a similar direction [11, 26].

Our system concept was closely associated with a

corresponding photonic integration concept, which was based

on the use of a low-cost polymer platform for the physical and

functional integration of a large number of passive and active

optical elements. The specific platform offers an extensive

toolbox of functionalities and ease of hybrid integration with

InP elements via low-loss butt coupling. Its high thermo-optic

coefficient (-110-4 K-1 to -310-4 K-1) and low thermal

conductivity (~0.3 W/m/K) allows the realization of highly

power efficient thermo-optic devices [27], while the ability to

integrate thin film elements inside trenches etched on the

platform, allows for on-chip polarization handling properties

like polarization rotation and polarization beam splitting or

combination [28]. In this first demonstration, single-flow (1-

flow) scenarios based on a dual-carrier or a dual-polarization

QPSK signal and 2-flow scenarios based on single

polarization QPSK signals were demonstrated at 28 Gbaud

without however any flexibility in the selection of the

modulation format or the symbol rate of each flow [24, 25].

In the present communication, we substantially extend our

previous work by demonstrating a fully flexible and

reconfigurable transmitter based on the same multi-flow

concept. The transmitter is capable of generating again up to

two optical flows, but this time with on-the fly selection of the

modulation format, symbol rate, wavelength allocation and

propagation direction per flow. The configuration of the

transmitter and the wavelength selective switch (WSS) inside

the optical part of the node is controlled by a software defined

optics (SDO) platform in real time. Our experimental

demonstration is realized at 12.5 and 25 Gbaud with single-

carrier, dual-carrier, single-polarization or dual-polarization

optical flows and with QPSK, 16-QAM and 64-QAM

modulation formats. Performance evaluation is successfully

carried out and demonstrated via bit-error rate measurements

after wavelength switching of the optical flows by the WSS

inside the optical node, and after subsequent transmission over

100 km of standard single-mode fiber (SSMF).

The remainder of the paper is organized as follows: Section

II describes the overall architecture and the design of the

transmitter, section III presents the development of the SDO

platform, and section IV presents the experimental setup and

the results. Finally, section V provides an outlook regarding

the integration of the transmitter, and gives the conclusions.

II. MULTI-FLOW TRANSMITTER CONCEPT AND

IMPLEMENTATION

Fig. 2 presents the layout of our transmitter concept elaborated

in detail in [24, 25], which comprises of two PolyBoard chips

and an array of four MZMs forming two IQMs (IQM1 and

IQM2). The back-end PolyBoard is used for the generation of

the optical carriers based on three hybridly integrated external

cavity lasers (ECLs) and for their optical routing towards the

MZM array inputs by means of the integrated thermo-optic

switches (TOS 1 and 2) that allow the light to pass either from

the upper or from the lower arm depending on their operation

state. The ECLs are based on the combination of InP gain

chips butt-coupled to the polymer chip which hosts three

tunable Bragg gratings and operate over 22nm in the C-band

[27-28]. The ECLs have a lasing threshold of ~5 mA, output

optical power higher than 5 dBm at 100 mA gain chip current

and 300 kHz linewidth. The tunable Bragg gratings consume

22 mW each, amounting to a tuning efficiency of 1 nm/mW.

The thermo-optic switches consist of simple Y-junctions with

off-set heater electrodes placed on each arm. Their typical

power consumption is 25 mW and the extinction ratio between

the two ports is higher than 20 dB. The front-end PolyBoard

is used for appropriately combining the outputs of the MZM

array to generate the different type of optical flows at the

transmitter outputs 1-3 considering polarization multiplexing

selectivity by means of the integrated polarization rotator (PR)

and the polarization beam combiner (PBC). The flexibility for

operation with either two flows or a single flow stems from

the possibility to operate independently the two IQMs or to

combine their outputs into a single optical entity. In the 2-flow

operation, the carriers from laser 1 and 3 are guided to IQM-1

and IQM-2. The corresponding modulation products appear at

points P1 and P2 and are routed to the output ports 1 and 3

corresponding to independent, single-carrier and single-

polarization signals. Depending on the analog driving signals

of each IQM, these modulation products correspond to simple

3

TOS 4

PBCPR

TOS 3

TOS 2

TOS 1

InP

InP

InP

2

1

PS

PS

MZM arrayBack-end polyboard

3

2

1

Front-end polyboard

IQM-2P2

IQM-1P1

Single-carrier, single-poloptical flow

Flow A:

Single-carrier, single-poloptical flow

Flow B:

Dual-carrier, single-poloptical flow

Flow C:

Single-carrier, dual-poloptical flow

Flow D:

2-flo

wo

peratio

n

1-flo

wo

pe

ration

1-flo

wo

pe

ration

Multi-flowTx 1

North

South

East

West

Clie

nt

dat

a

Data p

rocessin

g un

it

SDO platform

…… …

Multi-flowTx 2

1xN

WSS

1xM

WSS



3

(a)

(b)

Fig. 4: (a) Layout of the actual implementation of the multi-flow transmitter in this work with a back-end PolyBoard, two lithium niobate IQMs and a bulk

implementation of the front-end part. (b) Packaged back-end PolyBoard.

QPSK, 16-QAM or 64-QAM signals. In the 1-flow operation,

the products at P1 and P2 are combined either off-chip as a

dual-carrier signal or on-chip as a dual-polarization signal. In

the first case, the carriers are generated by laser 1 and 3 with

correlated wavelengths, and the modulation products appear at

ports 1 and 3. In the second case, a single carrier is generated

by laser 2. The modulation products are guided to the PBC

and appear at port 2.

Fig. 3 shows now the possible position of the transmitter

inside a network node and its possible interconnection with a

digital switch and a pair of WSSs. At the digital side, the

client data are organized in data flows that correspond to

independent end-to-end connections and feed the driving

circuits of each transmitter after proper selection of the

modulation format and symbol rate according to the flow size

and the corresponding transmission distance. At the optical

side, the output ports of each transmitter are connected to a

WSS, which is further connected back-to-back to a second

WSS for final routing. When the transmitter operates with two

optical flows (Flow A and B), these flows enter the first WSS

from different ports and are independently switched by the

second WSS. When on the other hand the transmitter operates

with one dual-carrier flow (Flow C), the two modulated

products enter the first WSS from different ports but are

switched as a single entity by the second WSS. Finally, when

the transmitter operates with one dual-polarization flow (Flow

D), this enters the first WSS from a single port and is switched

again to any direction by the second WSS. The SDO agent

that resides on top communicates with the digital and the

optical part of the transmitter and determines the number, the

type, the modulation format, the symbol rate, the wavelength

allocation and the switching direction of the optical flows.

Regarding the actual implementation of the multi-flow

transmitter, it is noted that the layout of Fig. 2 corresponds to

an ideal, fully integrated version. In this work, we use a

packaged polymer chip (PolyBoard) for the implementation of

the back-end part, two external lithium niobate IQMs, and a

bulk implementation of the front-end part with optical fibers

and bulk polarization controllers and PBC, as illustrated in

Fig. 4a. A picture of the packaged front-end PolyBoard is

given in Fig. 4b, and its design and characterization have been

previously presented in detail in [24]. It is noted that a

packaged front-end PolyBoard that was used in our proof-of-

concept demonstration in [24] was not available anymore

making necessary the use of bulk components for the

implementation of this part of the transmitter.

The modulation format and the symbol rate of each optical

flow depend on the number of levels and the rate of the multi-

level signals that feed the IQMs. These signals are generated

by the electrical driving elements, which in this work are

based on selector power digital-to-analog converters

(SPDACs) with 50 GHz bandwidth, fabricated in the indium

phosphide double heterojunction bipolar transistor (InP-

DHBT) technology [29]. Each SPDAC has 6 data inputs, 1

clock input and 2 outputs that provide complementary analog

signals with up to 8 levels and with amplitude swing up to 2 V

each. Each SPDAC combines in fact three different

functionalities, including 2:1 time division multiplexing, 3-bit

digital-to-analog conversion, and amplification. Depending on

the number of active data inputs (2, 4 or 6), each SPDAC can

provide an analog signal with 2, 4 or 8 levels and support (in

combination with another SPDAC) the operation of an IQM

with QPSK, 16-QAM or 64-QAM modulation format,

respectively. The selection of the modulation format of each

optical flow is thus associated with the encoding of the data

for each optical flow and the feeding of the SPDACs with the

proper number of input digital streams by the digital part of

the transmitter. In this work, the digital part is realized with

the help of a Field Programmable Gate Array (FPGA) board,

as it is explained in more detail in the next section.

III. MULTI-FLOW TRANSMITTER PROGRAMMABILITY

The SDO platform in this work provides the possibility for

controlling in an automated way and from a single user

interface the configuration of the optical and digital part of the

multi-flow transmitter, as well as the configuration of the

WSS inside the optical node. More specifically, our platform

communicates with the current sources that provide the

current to the InP gain chips, the current to the heaters of the

Bragg gratings, and the current to the TOSs of the back-end

PolyBoard. In this way, it can fully control the activation or

deactivation of the three tunable lasers, their emission

wavelength, and the state of the TOSs that ensures the

minimum optical loss on-board depending on the number and

type of the generated optical flows. It also communicates with

the FPGA board that generates, organizes and encodes the

data of the optical flows, and sends the digital streams and the

clock signals that feed the SPDACs of the two IQMs. Finally,

TOS1

TOS2

Back-end PolyBoard

PBCOUT2

OUT3

PC

PC

PDACSPDAC-3

S: Selector

TOS: Thermo-Optic Switch

IQM: IQ Modulator

SPDAC: Selector - power DAC

PC: Polarization controller

SP: Splitter

PBC: Polarization Beam Coupler

PDACSPDAC-4

PDACSPDAC-2

PDACSPDAC-1

IQM-2

IQM-1

FP

GA

SSS

SSS

SSS

SSS

OUT1

SP

SP



4

Fig. 5: Front panel of SDO platform for automated control and configuration

of the digital and optical part of the multi-flow transmitter. The inset shows

the drop-down menu for the selection of the operation case.

TABLE I

SUMMARY OF INVESTIGATED OPERATION CASES OF THE MULTI-FLOW

TRANSMITTER

Type of

operation

Case

no

IQM-1

format

IQM-2

format

IQM-1rate

(Gbaud)

IQM-2 rate

(Gbaud)

1-flow:

Dual Polarization

0 QPSK QPSK 25

1 16QAM 16QAM 12.5

2 16QAM 16QAM 25

3 64QAM 64QAM 25

1-flow:

Dual Carrier

4 QPSK QPSK 25 25

5 QPSK 64QAM 12.5 12.5

6 QPSK 64QAM 25 25

7 16QAM 16QAM 12.5 25

8 16QAM 16QAM 25 25

9 64QAM 64QAM 25 25

2-flow:

Single Polarization

Single Carrier

10 QPSK QPSK 25 25

11 QPSK 64QAM 25 25

12 16QAM 16QAM 12.5 25

13 16QAM 16QAM 25 25

14 64QAM 16QAM 25 25

15 64QAM 64QAM 25 25

it communicates with the WSSs of the node and controls the

spectral response of their ports in terms of central wavelength

and pass-band width, allowing for the formation of dual-

carrier signals (if this is the operation case), and for the

routing of the optical flows to their final destination.

Fig. 5 presents the user interface of the platform in LabVIEW.

The interface prompts the user to select the type of operation

(i.e. 1-flow or 2-flow operation), the type of the optical signal

in the case of 1-flow operation (i.e. dual-carrier or dual-

polarization), as well as the wavelength allocation, the output

port to the outer network, the symbol rate and the modulation

format for each optical flow. Based on this input, the tool

adjusts the current sources that control the elements on the

back-end PolyBoard as per the description above,

Fig. 6: Design concept for the driving of an IQM in our implementation using

the complementary outputs D and D/ of the FPGA transmitters and the complementary outputs of the SPDACs. The second IQM of the multi-flow

transmitter is driven in the same way.

Fig. 7: Examples of electrical signals at the output of the FPGA board for

driving one of the IQMs: (a) Example showing one active transmitter at 12.5 Gb/s and clock at 6.25 GHz for operation with QPSK format at 25 Gbaud, and

(b) example showing two active transmitters at 6.25 Gb/s and clock 3.125

GHz for operation with 16-QAM format at 12.5 Gbaud.

configures the two WSSs inside the optical node, and sends to

the FPGA board via a dedicated serial communication

interface (UART) a binary codeword, which is used to select

the appropriate FPGA state among a set of pre-defined

options. Within this context, the FPGA acts as a state machine

that configures its state in terms of number and bit rate of the

generated digital streams, according to the operation case of

the multi-flow transmitter that has been selected by the user

through the interface of the SDO platform.

Table I summarizes the operation cases that are considered

in this work and are associated with specific designs of the

FPGA, covering a very broad range of combinations regarding

the number and type of optical flows, as well as the

modulation format and the symbol rate per flow. It is noted

that Table I does not include the wavelength allocation and the

output direction of each flow, which are two additional

parameters that increases even further the diversity of the

investigated cases. Some of the cases in Table I are simpler

than others like for example the 2-flow cases, where the two

signals have the same modulation format and rate or the 1-

flow cases, where the two carriers of a dual-carrier signal have

again the same format and rate. On the other hand, some other

cases are more complex in the sense that either the format or

the rate are different among the two signals in 2-flow

operation (see cases 11-12) or among the two carriers in 1-

flow operation (see cases 5-7). Given the options in this work

for the modulation format (QPSK, 16-QAM or 64-QAM) and

the symbol rate (12.5 or 25 Gbaud), the maximum capacity of

FPGA

Tx: FPGA transmitter

D: Data

PS: Phase shifter

DL: Delay Line

FD: Frequency Doubler

Clk: Clock signal

90o

IQM-1

SPDAC-1

DL

90o

IQM-2

PS

FD

Clk-2

PRBS11101010

PDAC

SS S

PS PS

PRBS11

PS PS

PRBS11

D D/Tx-8

DL DL DL

SPDAC-2

DL

PSPS

D D/Tx-7

D D/Tx-6

D D/Tx-5

PSPS

FD

Clk-1

PRBS11101010

PDAC

SS S

PS PS

PRBS11

PS PS

PRBS11

D D/Tx-4

DL DL DL

D D/Tx-3

D D/Tx-2

D D/Tx-1

PS

PSPS

PSPS

12.5 Gb/s 6.25 Gb/s



5

Fig. 8: Experimental setup for the system evaluation of our multi-flow transmitter after transmission over 100 km of SSMF. The evaluation is made with respect

to the 16 operation cases of the transmitter summarized in Table I.

the transmitter is 300 Gb/s, and is fully used in the cases 3, 9

and 15. All other cases represent configurations that waste part

of the available capacity. However, they can be still

meaningful in a true networking scenario involving

temporarily low traffic and long or noisy links. It should be

also noted that the total capacity of the transmitter is limited in

this work by the performance of our lithium niobate IQMs.

Given the high bandwidth of the SPDACs, the symbol rate can

be extended to the 50 Gbaud regime, if high-speed modulators

based on InP [30] or electro-optic polymers [31-33] are

available, allowing for a corresponding capacity extension to

600 Gb/s.

Given the set of operation cases in Table I, each corresponding

FPGA design is associated with the generation of the proper

number of binary sequences at the proper rate, and the

activation of the proper number of FPGA transmitters in order

to feed the SPDACs. For example, for an IQM operating with

QPSK format at 25 Gbaud, the FPGA shouldprovide each

SPDAC of this IQM with 2 digital streams at 12.5 Gb/s and a

clock at 12.5 GHz. The latter can be generated as a 12.5 Gb/s

signal with alternating “1s” and “0s” (i.e. a clock at 6.25 GHz)

that passes through an external frequency doubler (FD). In a

different case of IQM operation with 16- QAM at 12.5 Gbaud,

the FPGA should provide each SPDAC of this IQM with 4

digital streams at 6.25 Gb/s and a clock signal at 6.25 GHz,

which can be generated again in a similar way with an initial

3.125 GHz clock and frequency doubling. Finally, in the case

of IQM operation with 64-QAM, the FPGA should provide

each SPDAC with 6 digital streams in order to have at the end

an 8-level driving signal at the final symbol rate. With

extension of this thinking to both IQMs, it can be easily found,

which number and what kind of binary streams should be

generated by the respective FPGA design for each case of

Table I. It is noted that in our implementation all binary

streams are generated by the FPGA board based on the same

pseudorandom bit sequence (PRBS) with length 211-1. Thus,

decorrelation between the streams that feed the different input

ports of the SPDACs is necessary and can be realized in the

digital domain on the FPGA board. It is also noted that the

number of FPGA transmitters in our implementation is smaller

than the number in a real system, due to the use of the

complementary outputs of a single transmitter at both input

ports of each selector in the SPDACs, and due to the use of the

complementary outputs of a single SPDAC for driving both

phase components (I and Q) of the IQMs. It becomes thus

clear that the FPGA board generates one clock signal and one,

two or three binary streams for each IQM corresponding to

QPSK, 16-QAM or 64-QAM operation, respectively. In order

to have this simplification in the experimental part, but allow

at the same time for pattern decorrelation and alignment at the

bit/symbol level, external microwave delay lines (DL) and

phase shifters (PS) are used, as shown in Fig. 6. As example,

Fig. 7 presents the electrical signals at the output of the FPGA

board that drive one of the IQMs in the case of QPSK

operation at 25 Gbaud and 16-QAM operation at 12.5 Gbaud.

IV. EXPERIMENTAL SETUP AND RESULTS

Fig. 8 illustrates the deployed experimental set up for the

assessment of the multi-flow transmitter. A Xilinx Virtex 7

Series FPGA evaluation board is used to generate the binary

streams and the corresponding clock signals, feeding the two

SPDACs, according to the analysis of section III. Given the

selected operation case, one or two optical carriers are

generated by the back-end part and feed the two LiNbO3

single polarization IQMs, after proper amplification and

adjustment of their polarization state. The IQMs exhibit 28

GHz 3-dB bandwidth while the required voltage for pi-shift is

3.5 V. Subsequently, the modulated signals enter the bulk

implementation of the front-end part. An optical delay line

(ODL) is used at the output of the upper IQM in order to



6

Fig. 9: Optical spectra of different types of operation with 100 Gb/s total capacity, corresponding to: a) 1-flow dual-polarization (case 0), b) 1-flow

dual-polarization with different format and rate (case 1), c) 1-flow dual-carrier

(case 4), and d) 2-flow single-polarization single-carrier operation (case 10).

achieve synchronization at the bit level in the case of dual

polarization operation. A pair of polarization controllers are

also used to ensure the required orthogonality between the

polarization states of the signals. The signals at the output of

the front-end part are combined by a 9x1 flex-grid WSS and

are wavelength switched by the second fixed grid 1x4 WSS to

the four possible directions of the ingress node. The WSSs are

based on LCoS technology and exhibit reconfiguration times

in the 10-100 ms range [34].

In the specific experimental setup, the optical flows are

transmitted to the north direction and are dropped at the egress

node after transmission over 100 km of SSMF. It is noted that

the performance of the generated optical flows is not affected

by the selection of the output WSS direction (i.e. output WSS

port). The egress node is emulated by an optical tunable filter

with sharp pass-band and variable width. The signals are

detected using a coherent optical receiver with polarization

diversity and 45 GHz 3-dB bandwidth. A low linewidth laser

(<100 kHz), tuned at the wavelength of the modulated signal,

serves as local oscillator ensuring minimization of the phase

noise and the frequency offset, respectively. Subsequently, the

four output electrical signals, corresponding to the in-phase

and quadrature components of the two polarization states, feed

a real-time oscilloscope, which exhibits 33 GHz 3-dB

bandwidth (Agilent DSAX93304Q), where are sampled and

stored for offline digital signal processing (DSP).

Fig. 9 depicts indicative optical spectra at the north output

of the 1x4 WSS in the case of the three types of operation (i.e.

1-flow dual-polarization, 1-flow dual-carrier, 2-flow single-

polarization, single-carrier) generating a total capacity of 100

Gb/s per flow. For 1-flow dual-polarization operation the

wavelength λ1 is set at 1551.65 nm, while for 1-flow dual-

carrier operation, we use a 100 GHz spacing between the

carriers, setting the wavelengths at λ1 and λ1-0.8 nm (1550.85

nm) due to the limitation of using the fixed grid WSS at the

Fig. 10: Evaluation in b2b and after 100 km transmission of different types of

operation with 100 Gb/s total capacity, corresponding to 1-flow dual-

polarization (case 0), 1-flow dual-carrier (case 4 and 5), and 2-flow single-

polarization single-carrier operation (case 10). The eye-diagrams correspond

to b2b for case 0 (upper row), case 4 (middle row) and case 5 (lower row).



polarization (case 2), 1-flow dual-carrier (case 6 and 8), and 2-flow single-polarization single-carrier operation (case 11 and 13). The constellation

diagrams correspond to b2b for case 2 (upper row), case 6 (middle row) and

case 13 (lower row).

output of the ingress node. It is noted that in the case of using

two FlexGrid WSS instead, the two optical carriers can be

spaced at frequency separation of multiples of 12.5 GHz and

as narrow as the bandwidth of the modulated signals. In the

c)

a) b)

1540 1544 1548 1552 1556 1560

-70

-60

-50

-40

-30

-20

Level(dBm)

WaveLength(nm)

case 4-DC QPSK-25G

λ1-0.8 nm λ1

1540 1544 1548 1552 1556 1560

-70

-60

-50

-40

-30

-20Level(dBm)

WaveLength(nm)

case 2-DP-16QAM 25G

λ1

1540 1544 1548 1552 1556 1560

-70

-60

-50

-40

-30

-20

Level(dBm)

WaveLength(nm)

case 1-DP 16QAM-12.5G

λ1

Flow A

Flow Cd)

Case 4: 1-flow DC SP-QPSK-25G

Case 0: 1-flow DP-QPSK-25G Case 1: 1-flow DP-16QAM-12.5G

1540 1544 1548 1552 1556 1560

-70

-60

-50

-40

-30

-20Level(dBm)

WaveLength(nm)

case D-16QAM 25G-16QAM 25G

λ2 λ1

Case 10: 2-flow SP-16QAM-25G

Flow D+E

Flow B

DP-QPSK 25

GBd

SP-QPSK 25 GBd

SP-QPSK 12.5 GBd & SP-64QAM 12.5 GBd

-22 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2

0

2

4

6

8

10

12

14

16

18

20

22

Q f

ac

tor

(dB

)

Received Optical Power (dBm)

Case 0-DP-QPSK 25G

Case 0-DP-QPSK 25G 100km

Case 4 IQM1-SP-QPSK 25G

Case 4 IQM1-SP-QPSK 25G 100km



Case 5 IQM1-SP-QPSK 12.5G

Case 5 IQM1-SP-QPSK 12.5G 100km

case 5 - IQM2-SP-64QAM 12.5G

case 5 - IQM2-SP-64QAM 12.5G 100km





FEC-OH 24%

FEC-OH 7%

case 0 – DP-QPSK 25GBdcase 0 – DP-QPSK 25GBd 100kmcase 4 – IQM1-SP-QPSK 25GBdcase 4 – IQM1-SP-QPSK 25GBd 100km

case 5 – IQM1-SP-QPSK 12.5GBdcase 5 – IQM1-SP-QPSK 12.5GBd 100km

case 4 – IQM2-SP-QPSK 25GBdcase 4 – IQM2-SP-QPSK 25GBd 100km

case 5 – IQM2-SP-64QAM 12.5GBdcase 5 – IQM2-SP-64QAM 12.5GBd 100kmcase 10 – IQM1-SP-QPSK 25GBdcase 10 – IQM1-SP-QPSK 25GBd 100kmcase 10 – IQM2-SP-QPSK 25GBdcase 10 – IQM2-SP-QPSK 25GBd 100km

IQM1 IQM2

IQM1 IQM2

1-flow Dual Carrier

1-flow Dual Polarization

1-flow Dual Carrier

Qu

ad

ratu

re

Qu

ad

ratu

re

In PhaseIn Phase

X pol. Y pol.

DP-16QAM 25 GBd

SP-16QAM 25 GBd

IQM1 IQM2

Qu

ad

ratu

re

Qu

ad

ratu

re

In PhaseIn Phase

SP-QPSK 25 GBd & SP-64QAM 25 GBd

Qu

ad

ratu

re

Qu

ad

ratu

re

In PhaseIn Phase

IQM1 IQM2

-22 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2

0

2

4

6

8

10

12

14

16

18

20

22

Q f

ac

tor

(dB

)


case 2-DP-16QAM 25G

case 2-DP-16QAM 25G 100km

use case 6 - IQM1-SP-QPSK 25G

use case 6 - IQM1-SP-QPSK 25G 100km

use case 6 - IQM2-SP-64QAM 25G

use case 6 - IQM2-SP-64QAM 25G 100km

case 8 IQM1-SP-16QAM 25G

case 8 IQM1-SP-16QAM 25G 100km





Case 11 IQM2-SP-64QAM 25G

Case 11 IQM2-SP-64QAM 25G 100km





FEC-OH 24%

FEC-OH 7%

case 2 – DP-16QAM 25GBdcase 2 – DP-16QAM 25GBd 100kmcase 6 – IQM1-SP-QPSK 25GBdcase 6 – IQM1-SP-QPSK 25GBd 100km

case 8 – IQM1-SP-16QAM 25GBdcase 8 – IQM1-SP-16QAM 25GBd 100km



case 13 – IQM1-SP-16QAM 25GBdcase 13 – IQM1-SP-16QAM 25GBd 100kmcase 13 – IQM2-SP-16QAM 25GBdcase 13 – IQM2-SP-16QAM 25GBd 100km

case 11 – IQM1-SP-QPSK 25GBdcase 11 – IQM1-SP-QPSK 25GBd 100kmcase 11 – IQM2-SP-64QAM 25GBdcase 11 – IQM2-SP-64QAM 25GBd 100km

2-flow Single Polarization Single Carrier

1-flow Dual Polarization

1-flow Dual Carrier



7



polarization (case 3), 1-flow dual-carrier (case 9), and 2-flow single-

polarization single-carrier operation (case 15). The constellation diagrams and

the eye-diagram correspond to b2b for case 3 (upper row) and case 9 (lower

row).

Fig. 13: Picture of test subassembly enabling the interconnection of 4

SPDACs with an InP chip via an RF interposer. This subassembly will be

further integrated with a back-end and a front-end PolyBoard and will be part

of our multi-flow transmitter, integrated as a single, small-form factor device.

case of two independent flows, the first flow is centered at λ1,

while the second one at λ2 (1542.75 nm) with 1 THz spacing,

to showcase the wavelength insensitivity of the transmitter.

The transition between the three types is controlled by the

SDO agent and takes place within 50-100 ms being limited

again by the response of the thermo-optic elements on the

PolyBoard and the reconfiguration time of the WSSs [24, 27].

The SDO communication interface with the optical hardware

elements is very fast and takes place within 1-3.5 ms which is

negligible to the overall reconfiguration time of the multi-flow

transmitter. The assessment of the transmitter is based on the

calculation of the Q-factor of the detected signals as a function

of the optical power at the input of the coherent receiver. The

results have been grouped into three data sets according to the

total capacity of the corresponding use case. Fig. 10 presents

the Q-factor curves for the 1-flow and 2-flow cases 0, 4, 5, 10,

where each flow has a total capacity of 100 Gb/s. Error free

operation well above the FEC limit with 7% overhead (BER =

3.8E-3) is achieved for the flows with QPSK modulation

format at 25 Gbaud for both back-to-back (b2b) configuration

and transmission after 100 km. The performance of the 64-

QAM signals at 12.5 Gbaud is limited by the low voltage

swing of the electrical signals that feed the IQ modulators.

However, for optical power higher than -6 dBm, the

corresponding Q-factor values are above the FEC limit with

24% overhead (BER = 4.5E-2). Indicative eye-diagrams from

the three operation cases are also presented in fig. 10.

The second data set includes the cases 2, 6, 8 ,11 and 13,

where the total capacity is 200 Gb/s. Fig. 11 illustrates the

corresponding Q-factor curves against the received optical

power and indicative constellation diagrams for 0 dBm

received power in b2b configuration. The single-polarization

QPSK and 16-QAM, as well as the DP-QPSK signals have a

Q-factor above the FEC limit with 7% overhead for received

power higher than -8 dBm, while the single polarization 64-

QAM signals are above the FEC limit with 24% overhead for

optical power above -4 dBm.

Finally, the third data set includes the cases 3, 9 and 15

where the total capacity is 300 Gb/s. Fig. 12 presents the

corresponding evaluation results for b2b and transmission

after 100 km. The first curve corresponds to the dual-

polarization flow with 64QAM at 25 Gbaud (case 3), the

second one to the dual-carrier flow with two SP-64QAM

signals at 25 Gbaud (case 9), and the third one to the case of

two independent flows, each with SP-64QAM format at 25

Gbaud (case 15). In all cases Q-factor performance above the

FEC limit with 24% overhead is obtained for received power

higher than -4 dBm.

V. CONCLUSIONS AND NEXT STEPS

We have extended in this work our proof-of-concept

demonstration of a multi-flow transmitter based on the

combination of spectrum and polarization sliceability, and the

use of photonic integration on the PolyBoard platform [24].

The extension in this work consists in the use of two 3-bit

SPDACs as the driving elements of the IQMs of the

transmitter, and the development of a practical SDO platform

that enables the configuration of the transmitter in terms of

number, type, emission wavelength, modulation format,

symbol rate and output direction of the optical flows. We have

used as example a set of 16 different configuration cases

involving operation with QPSK, 16-QAM or 64-QAM at 12.5

or 25 Gbaud, and representing different combinations of the

number, type, format and rate of the optical flows with total

capacity up to 300 Gb/s. Using this transmitter inside an

optical node and making coherent transmission experiments

over 100 km of SSMF, we have demonstrated flexible

operation with interchange between the different operation

cases and sufficient transmission performance with Q-factor

Qu

ad

ratu

re

Qu

ad

ratu

re

In PhaseIn Phase

X pol. Y pol.

-22 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2

0

2

4

6

8

10

12

14

16

Q f

ac

tor (

dB

)


Case 3-DP-64QAM 25G

Case 3-DP-64QAM 25G 100km









FEC-OH 24% Qu

ad

ratu

re

Qu

ad

ratu

re

In PhaseIn Phase

IQM1 IQM2

IQM1 IQM2

FEC-OH 7%SP-64QAM 25 GBd

2-flow Single Polarization Single Carrier

1-flow Dual PolarizationDP-64QAM 25 GBd

case 3 – DP-64QAM 25GBdcase 3 – DP-64QAM 25GBd 100kmcase 9 – IQM1-SP-64QAM 25GBdcase 9 – IQM1-SP-64QAM 25GBd 100km






8

values below the FEC limit in all cases.

Next steps in this work will evolve along two different

directions. The first one is associated with the integration of

the multi-flow transmitter as a single, small-form factor device

based on the integration of an InP MZM chip with a back-end

and a front-end PolyBoard, as per the diagram in Fig. 2.

Progress on the design of a radio-frequency (RF) interposer

that will enable the interconnection of the SPDACs with the 4

MZMs in order to feed the high-speed RF data streams at the

output of the driver ICs to the inputs of the modulators

synchronized and with minimum loss, and progress on the

design of a method for attaching this interposer on the top of

the InP chip have been already good and have led to compact

subassembly structures, as shown in Fig. 13. The second

direction involves improvements in our SDO platform in order

to perform true traffic aggregation tasks, collecting Ethernet or

Fibre Channel traffic, and organizing this traffic into flows

that will be transmitted via Optical Transport Network (OTN)

frames. Furthermore, work on the development of the

appropriate Yang models and the necessary extensions of a

southbound SDN protocol e.g. OpenFlow [35] will be carried

out in order to integrate the flexible multi-flow transmitter and

WSS elements to an SDN platform like ONOS [36]. In this

way, abstraction of the reconfigurable optical transport

parameters will be possible to an SDN controller allowing full

network programmability.

ACKNOWLEDGEMENT

The present work was supported by the European Commission

through the project ICT-PANTHER (Contract number

619411) funded under the 7th Framework Programme (FP7).

REFERENCES

[1] Weldon, M. (2015). The Future X Network. Boca Raton: CRC Press. [2] Global - 2020 Forecast Highlights – Cisco,

https://www.cisco.com/c/dam/m/en_us/solutions/service-provider/vni-

forecast-highlights/pdf/Global_2020_Forecast_Highlights.pdf [3] K. Roberts, Q. Zhuge, I. Monga, S. Gareau and C. Laperle, “Beyond 100

Gb/s: capacity, flexibility, and network optimization [Invited],” in IEEE/OSA Journal of Optical Communications and Networking, vol. 9,

no. 4, pp. C12-C23, April 2017.

[4] M. Jinno, "Elastic Optical Networking: Roles and Benefits in Beyond 100-Gb/s Era," in Journal of Lightwave Technology, vol. 35, no. 5, pp.

1116-1124, March1, 1 2017.

[5] H. Takara, K. Yonenaga and M. Jinno, “Spectrally-efficient Elastic Optical Path Networks Toward 1 Tbps Era,” in Proc. OFC/NFOEC

2012, Los Angeles CA, Mar. 2012, paper OTh3B.3.

[6] A. dupas, P. layec, E. dutisseuil, S. bigo, S. belotti, S. misto, S. annoni, Y. yan, E. Hugues Salas, G. S. Zervas, and D. Simeonidou, "Hitless 100

Gbit/s OTN Bandwidth Variable Transmitter for Software-Defined

Networks," in Optical Fiber Communication Conference, OSA Technical Digest (online) (Optical Society of America, 2016), paper

Th3I.1.

[7] Yan, S., Hugues-Salas, E., Hammad, A., Yan, Y., Saridis, G., Bidkar, S., Nejabati, R., Simeonidou, D., Dupas, A. & Layec, “Demonstration of

Bandwidth Maximization between Flexi/Fixed Grid Optical Networks

with Real-Time BVTs,” ECOC 2016; 42nd European Conference on Optical Communication, Dusseldorf, Germany, 2016, pp. 1-3.

[8] J. M. Fàbrega, M. S. Moreolo, A. Mayoral, R. Vilalta, R. Casellas, R.

Martinez, R. Muñoz, Y. Yoshida, K. Kitayama, Y. Kai, M. Nishihara, R. Okabe, T. Tanaka, T. Takahara, J. C. Rasmussen, N. Yoshikane, X. Cao,

T. Tsuritani, I. Morita, K. Habel, R. Freund, V. López, A. Aguado, S.

Yan, D. Simeonidou, T. Szyrkowiec, A. Autenrieth, M. Shiraiwa, Y. Awaji, and N. Wada, "Demonstration of Adaptive SDN Orchestration: A

Real-Time Congestion-Aware Services Provisioning Over OFDM-Based 400G OPS and Flexi-WDM OCS," in Journal of Lightwave Technology,

vol. 35, no. 3, pp. 506-512, Feb.1, 1 2017.

[9] S. Yan, Y. Yan, B. R. Rofoee, Y. Shu, E. Hugues-Salas, G. Zervas, and D. Simeonidou "Real-Time Ethernet to Software-Defined Sliceable

Superchannel Transponder," in Journal of Lightwave Technology, vol.

33, no. 8, pp. 1571-1577, April15, 15 2015. [10] Y. Ou, S. Yan, A. Hammad, B. Guo, S. Peng, R. Nejabati, and D.

Simeonidou, "Demonstration of Virtualizeable and Software-Defined

Optical Transceiver," in Journal of Lightwave Technology, vol. 34, no. 8, pp. 1916-1924, April 15, 15 2016.

[11] M. S. Moreolo, J. M. Fàbrega, L. Nadal, F. J. Vílchez, A. Mayoral, R.

Vilalta, R. Muñoz, R. Casellas, R. Martinez, M. Nishihara, T. Tanaka, T. Takahara, J. C. Rasmussen, C. Kottke, M. Schlosser, R. Freund, F.

Meng, S. Yan, G. Zervas, D. Simeonidou, Y. Yoshida, and K.-I.

Kitayama, "SDN-Enabled Sliceable BVT Based on Multicarrier Technology for Multiflow Rate/Distance and Grid Adaptation," in

Journal of Lightwave Technology, vol. 34, no. 6, pp. 1516-1522, March

15, 2016. [12] B. G. Saavedra et al., “Multi-channel Mach-Zehnder IQ modulator PICs

on InP for hybrid OFDM transmitter integration,” 2016 18th

International Conference on Transparent Optical Networks (ICTON),

Trento, 2016, pp. 1-4.

[13] M. Seimetz, “Multi-format transmitters for coherent optical M-PSK and

M-QAM transmission,” in Proceedings of 2005 7th International Conference Transparent Optical Networks, 2005, pp. 225-229, Vol. 2.

[14] K. Yonenaga, F. Inuzuka, S. Yamamoto, H. Takara, B. Kozicki, T. Yoshimatsu, A. Takada, Masahiko Jinno, "Bit-rate-flexible all-optical

OFDM transceiver using variable multi-carrier source and

DQPSK/DPSK mixed multiplexing," 2009 Conference on Optical Fiber Communication - incudes post deadline papers, San Diego, CA, 2009,

pp. 1-3.

[15] N. Sambo et al., “Next generation sliceable bandwidth variable transponders,” in IEEE Communications Magazine, vol. 53, no. 2, pp.

163-171, Feb. 2015.

[16] M. Jinno, H. Takara, Y. Sone, K. Yonenaga and A. Hirano, "Multiflow optical transponder for efficient multilayer optical networking," in IEEE

Communications Magazine, vol. 50, no. 5, pp. 56-65, May 2012.

[17] H. Takara, T. Goh, K. Shibahara, K. Yonenaga, S. Kawai and M. Jinno, "Experimental demonstration of 400 Gb/s multi-flow, multi-rate, multi-

reach optical transmitter for efficient elastic spectral routing," 2011 37th

European Conference and Exhibition on Optical Communication, Geneva, 2011, pp. 1-3.

[18] A. Pagano, E. Riccardi, M. Bertolini, V. Farelli and T. Van De Velde,

"400Gb/s real-time trial using rate-adaptive transponders for next generation flexible-grid networks," OFC 2014, San Francisco, CA,

2014, pp. 1-3.

[19] T. Tanaka, A. Hirano and M. Jinno, "Impact of Transponder Architecture on the Scalability of Optical Nodes in Elastic Optical

Networks," in IEEE Communications Letters, vol. 17, no. 9, pp. 1846-

1848, September 2013. [20] T. Tanaka, A. Hirano, and M. Jinno, "Advantages of IP over elastic

optical networks using multi-flow transponders from cost and equipment

count aspects," Opt. Express 22, 62-70 (2014). [21] M. Jinno, Y. Sone, H. Takara, A. Hirano, K. Yonenaga and S. Kawai,

"IP traffic offloading to elastic optical layer using multi-flow optical

transponder," 2011 37th European Conference and Exhibition on Optical Communication, Geneva, 2011, pp. 1-3.

[22] Thierry Zami, "Multiflow Application for WDM Networks With

Multicarrier Transponders Serving Superchannels in Contentionless OXCs [Invited]," J. Opt. Commun. Netw. 9, A114-A124 (2017).

[23] T. Tanaka, A. Hirano and M. Jinno, "Impact of multi-flow transponder

on equipment requirements in IP over elastic optical networks," 39th European Conference and Exhibition on Optical Communication (ECOC

2013), London, 2013, pp. 1-3.

[24] V. Katopodis, D. Felipe, C. Tsokos, P. Groumas, M. Spyropoulou, A. Beretta, A. Dede, M. Quagliotti, A. Pagano, A. Vannucci, N. Keil, H.

Avramopoulos, and Ch. Kouloumentas, "Multi-Flow Transmitter Based

on Polarization and Optical Carrier Management on Optical Polymers," in IEEE Photonics Technology Letters, vol. 28, no. 11, pp. 1169-1172,

June1, 1 2016.

[25] V. Katopodis, M. Spyropoulou, C. Tsokos, P. Groumas, D. Felipe, N. Keil, A. Beretta, A. Vannucci, T. K. Johansen, M. Quagliotti, A. Pagano,

J.-Y. Dupuy, A. Konczykowska, C. Delezoide, H. Mardoyan, Ch.

Kouloumentas and H. Avramopoulos, "Polarization-, carrier-, and

https://www.cisco.com/c/dam/m/en_us/solutions/service-provider/vni-forecast-highlights/pdf/Global_2020_Forecast_Highlights.pdf

https://www.cisco.com/c/dam/m/en_us/solutions/service-provider/vni-forecast-highlights/pdf/Global_2020_Forecast_Highlights.pdf



9

format-selectable optical flow generation based on a multi-flow transmitter using passive polymers," 2016 18th International Conference

on Transparent Optical Networks (ICTON), Trento, 2016, pp. 1-4.

[26] L. Nadal, M. Svaluto Moreolo, J. M. Fabrega and F. J. Vilchez, "Experimental Demonstration of a Programmable S-BVT with PDM

Capability for Flexible Optical Metro Networks," ECOC 2016; 42nd

European Conference on Optical Communication, Dusseldorf, Germany, 2016, pp. 1-3.

[27] Z. Zhang, N. Keil. (2015, August). Thermo-optic devices on polymer

platform. Optics Communications. [Online]. 362, 101-114. Available: http://dx.doi.org/10.1016/j.optcom.2015.08.026.

[28] N. Keil et al., "Novel integration technologies for disruptive capacity

upgrade in data center systems," 2016 Optical Fiber Communications Conference and Exhibition (OFC), Anaheim, CA, 2016, pp. 1-3.

[29] A. Konczykowska, J. Dupuy, F. Jorge, M. Riet, and V. Nodjiadjim,

“Extreme Speed Power-DAC: Leveraging InP DHBT for Ultimate Capacity Single-Carrier Optical Transmissions,” in Optical Fiber

Communication Conference, OSA Technical Digest (online) (Optical

Society of America, 2017), paper M3D.4. [30] E. Rouvalis et al., "A low insertion loss and low Vπ InP IQ modulator

for advanced modulation formats," 2014 The European Conference on

Optical Communication (ECOC), Cannes, 2014, pp. 1-3.

[31] V. Katopodis, C. Kouloumentas, A. Konczykowska, F. Jorge, P.

Groumas, Z. Zhang, A. Beretta, A. Dede, J.-Y. Dupuy, V. Nodjiadjim,

G. Cangini, G. Von Büren, E. Miller, R. Dinu, J. Han Choi, D. Pech, N. Keil, H.-G. Bach, N. Grote, A. Vannucci, and H. Avramopoulos, “Serial

100 Gb/s connectivity based on polymer photonics and InP-DHBT electronics,” Opt. Express, vol. 20(8), pp. 28538-28543, Dec. 2012.

[32] P. Groumas, Z. Zhang, V. Katopodis, A. Konczykowska, J. Y. Dupuy,

A. Beretta, A. Dede, J. H. Choi, P. Harati, F. Jorge, V. Nodjiadjim, M. Riet, R. Dinu, G. Cangini, E. Miller, A. Vannucci, N. Keil, H. G. Bach,

N. Grote, M. Spyropoulou, H. Avramopoulos and C. Kouloumentas

"Tunable 100 Gbaud Transmitter Based on Hybrid Polymer-to-Polymer Integration for Flexible Optical Interconnects," in Journal of Lightwave

Technology, vol. 34, no. 2, pp. 407-418, Jan. 2016.

[33] V. Katopodis, P. Groumas, Z. Zhang, R. Dinu, E. Miller, A. Konczykowska, J.-Y. Dupuy, A. Beretta, A. Dede, J.H. Choi, P. Harati,

F. Jorge, V. Nodjiadjim, Muriel Riet, G. Cangini, A. Vannucci, N. Keil,

H.-G. Bach, N. Grote, H. Avramopoulos, Ch. Kouloumentas, “Polymer enabled 100Gbaud connectivity for datacom applications,” Opt.

Commun., Vol. 362, 2016, pp. 13-21,

[34] M. Wang, L. Zong, L. Mao, A. Marquez, Y. Ye, H. Zhao, F.J. Vaquero Caballero, “LCoS SLM Study and Its Application in Wavelength

Selective Switch,” Photonics, Vol. 4, No. 2, 2017.

[35] S. Yamashita, A. Yamada, K. Nakatsugawa, T. Soumiya, M. Miyabe and T. Katagiri, "Extension of OpenFlow protocol to support optical

transport network, and its implementation," 2015 IEEE Conference on

Standards for Communications and Networking (CSCN), Tokyo, 2015, pp. 263-268.

[36] https://onosproject.org/

multi-flow transmitter with full format and rate ... · additional number of optical carriers is...

Documents